<i>in silico</i> Kv2.1 is widely expressed in brain, and inhibiting Kv2.1 is a potential strategy to prevent cell death and achieve neuroprotection in ischemic stroke. Herein, an model of Kv2.1 tetramer structure was constructed by employing the AlphaFold‐Multimer deep learning method to facilitate the rational discovery of Kv2.1 inhibitors. GaMD was utilized to create an ion transporting trajectory, which was analyzed with HMM to generate multiple representative receptor conformations. The binding site of RY785 and RY796(S) und‐ er the P‐loop was defined with Fpocket program together with the competitive binding electrophysiology assay. The docking poses of the two inhibitors were predicted with the aid of the semi‐empirical quantum mechanical calculation, and the IGMH results suggested that Met375, Thr376, and Thr377 of the P‐helix and Ile405 of the S6 segment made significant contributions to the binding affinity. These results provided insights for rational molecular design to develop novel Kv2.1 inhibitors.